CROSS-BORDER MERGERS
AS INSTRUMENTS OF
COMPARATIVE ADVANTAGE
J. Peter Neary
University College Dublin and CEPR
www.ucd.ie/~economic/staff/pneary/neary.htm
5. Cross-Border Mergers and Acquisitions
So far: Greenfield FDI only
BUT: Cross-border M&As are quantitatively much
more important
Now: Oligopoly model essential (almost)
•
•
•
No: Barba Navaretti/Venables (2003), Nocke/Yeaple (2004), Head/Ries (2005)
Yes: Long/Vousden (RIE 1995), Falvey (WE 1998), Horn/Persson (JIE 2001), etc.
Here: Neary (2004)
Model of 2-country integrated market:
• Cournot oligopoly
• Home: n firms with cost c; Foreign: n* with cost c*
• Absent mergers: “Cone of diversification” in {c, c*} space
2
Cross-border mergers:
• M&A’s a huge % of all FDI: more than greenfield FDI
• A high % of mergers are cross-border
• Cross-border merger waves linked to market integration
[EU Single Market; Mercosur]
How to explain them?
• I.O.:
• Strategic and efficiency motives
• All partial equilibrium
• Macro: Major innovations
Jovanovic/Rousseau (2003)
• International Trade Theory:
• Dominant paradigms: Competition (perfect/monopolistic)
• Needed: Oligopoly in GE
3
Plan
1.
2.
3.
4.
Specialisation Patterns in the Absence of Mergers
Myopic Mergers
Forward-Looking Mergers
General Oligopolistic Equilibrium:
•
•
Factor and Goods Markets: Ricardo + Cournot
Demand: Continuum-Quadratic Preferences
5. Mergers in General Equilibrium
6. Mergers and Welfare
4
1. Specialisation Patterns Without Mergers
•
•
•
•
•
•
•
Consider a typical sector, in partial equilibrium
Homogeneous-good Cournot competition
2 countries, integrated world market
Perceived linear demands: p = a  - b x
Given numbers of firms at home & abroad: n, n*
Firms in each country have identical costs: c, c*
Equilibrium home sales:
a '(n*  1)c  n*c*
y
b' (n  n*  1)
• So: y>0 
a '  n *c *
c *
n 1
 c  0a'(1  0 )c*
• Also holds with no foreign firms:
n*  0  0  1  c  a'
5
c
H firms unprofitable when n*=0
a'
a'
c*
6
c
a'
c  0a'(1  0 )c*
H firms unprofitable
when n*>0
a'
0a '  *
n 1
a'
c*
7
c
a'
H firms profitable
a'
n*  1
a'
c*
8
Symmetrically:
c
a'
F firms profitable
a'
n*  1
a'
n 1
a'
c*
9
c
F: Foreign
production only
O: No home or
foreign production
a'
a'
n*  1
HF: Home
and foreign
production
a'
n 1
H: Home
production only
a'
c*
Equilibrium Production Patterns
for Arbitrary Home and Foreign Costs
10
c
c~
F: Foreign
production only
O: No home or
foreign production
p(c,c*;n,n*)=0
HF: Home
and foreign
production
H: Home
production only
p*(c,c*;n,n*)=0
c~ *
c*
Fig. 4: Equilibrium Production Patterns
in Free Trade without FDI
11
Compare with perfect competition:
c
O: No home or
foreign production
a'
F: Foreign
production only
H: Home
production only
a'
c*
12
2. Myopic Mergers
Assumption 1: Only bilateral mergers can occur.
Immediate gain from a merger:
GFH (n, n* )  p * (n  1, n* )  p * (n, n* )  p (n, n* )
Assumption 2: A merger will not take place if GFH is zero
or negative.
Assumption 3: A merger will take place if GFH is strictly
positive.
13
Myopic Mergers (cont.)
What are the incentives to merge?
• No incentive if all firms are identical (and n+n*>2)
[Salant, Switzer, Reynolds (QJE 1983): “Cournot merger paradox”]
• Nor if the 2 merging firms are identical (and n+n*>2)
[Proposition 1]
• Intuition:
GHH (n, n* )  p (n  1, n* )  2p (n, n* )
• i.e., the profits of the acquiring firm would have to
double for such a merger to be profitable
• Same for all firms like the acquiring firm
14
Myopic Mergers (cont.)
• BUT: Outputs are strategic substitutes
15
Myopic Mergers (cont.)
• BUT: Outputs are strategic substitutes
y F
yF
• Removing a rival shifts down the
reaction functions of all others
• Movement along own reac. func.
• So: Outputs and operating profits
rise for all surviving firms
(including the acquiring firm)
• Hence: If firms differ in cost, a lowcost/high-cost takeover may be
profitable
16
Proposition 2: If c>c*, a takeover by a foreign firm is
profitable if the home firm has sufficiently high costs:
GFH>0 IFF:
c  1a '(1  1 )c*
0  1  0
Proof:
GFH  ( y1* ) 2  ( y0* ) 2  y02
 ( y1*  y0* )( y1*  y0* )  y02
Lemma: y1*  y0*  n1 y0
 n1 y0 ( y1*  y0*  ny0 )
 c  1a'(1  1 )c*
1  (   ) / 
  2n (n  1)  n* (n 2  2n  1)
    (n  2n  1)
2
   0
17
c
F
Incentives for
foreign firms to
take over home
O
H
0a'
1 a'
HF
c*
18
p
Dp*
GFH < 0
GFH > 0
Q
R
a–c*
a–c
Fig. 5: The Components of Gain
from a Cross-Border Acquisition by a Foreign Firm
19
Similarly, GHF >0 IFF:
c*  1*a '(1  1* )c
c
F
Incentives for
foreign firms to
take over home
O
H
0a'
1 a'
HF
Incentives for
home firms to
take over foreign
1*a '
c*
Fig. 2: Takeover Incentives
20
c
F
Incentives for
foreign firms to
take over home
O
H
p=0
Incentives for
home firms to
take over foreign
GFH=0
HF
GHF=0
c*
Fig. 6: Cross-Border Merger Incentives
21
Effects of one takeover on incentives for more?
• All surviving firms have higher output and profits:
Dy  D y* > 0
• Low-cost firms have higher output to begin with
• So, their profits rise by more: GFH is decreasing in n
[Proposition 3]
• i.e., “Merger Waves” / “Domino Mergers”:
=>
• No high-cost firms survive
• Mergers may not take place if n is large, even though
further mergers would be profitable
• Encouraging “national champions” by promoting
domestic mergers in high-cost sectors makes foreign
takeovers more likely (in the absence of cost synergies)
22
5. Cross-Border Mergers and Acquisitions (cont.)
Merger gains:
• For an acquisition of a home by a foreign firm:
GFH(c, c*; n, n*) = Dp*(.)  p(.)
• Always negative between identical firms
Salant/Switzer/Reynolds (QJE 1983) “Cournot merger paradox”
• Positive for a sufficiently large cost advantage
23
5. Cross-Border Mergers and Acquisitions (cont.)
So: Autarky to free trade encourages cross-border
M&As
Further results:
• GFH decreasing in n: Merger waves
• GFH decreasing in t (definitely for high t)
So partial trade liberalisation encourages cross-border M&As
Empirical evidence:
• Brakman/Garretsen/van Marrewijk (2005): Evidence in
favour of comparative advantage and merger waves
24
p
Dp*
Q
R R'
a–c*
a–c
Fig. 5a: Merger Waves:
Effects of a Fall in n
25
4. General Oligopolistic Equilibrium
• Continuum of sectors, indexed z  [0,1]
• Ricardian cost structure:
c(z) = wa(z),
c*(z) = w*a*(z)
• Assume home more efficient in low-z sectors
• 2 threshold sectors
[Perfect competition: c(z)=c*(z) is the threshold for specialisation]
26
c
Foreign
production only
c(1)
O
z 1
z~
z
c(0)
Home and
foreign
production
c*(1)
Home
production only
z~
z*
z0
c*(0)
c*
Fig. 1: Equilibrium Production Patterns
for a Given Cost Distribution
27
Demand
“Continuum-quadratic” preferences:

Max U [{x( z)}] 
subject to:


1
0
1
0
[ax( z)  21 bx( z) 2 ]dz
p( z) x( z)dz  I
p( z )   [a  bx ( z )],  
1
ap  bI
 p2
• Objective demand functions: depend on income and all prices
• Summing over 2 countries => Linear subjective demand funcs:

p( z)  a 'b' x ( z), a '  (a  a * ) /  , b'  b /  ,     *
28
GOLE: The Full Model
Three nominal variables: w, w*, 
• Absolute values are indeterminate
• Convenient normalisation: W=w, W*=w*
Full employment:
L  L(W ,W * , ~
z)
L*  L* (W ,W * , ~
z *)
Threshold sectors:
y( ~
z ,W ,W * )  0, ~
z 1
y* ( ~
z * ,W ,W * )  0,
~
z* 0
29
5. Mergers in General Equilibrium
Assume symmetric countries:
• n=n*, L=L*, etc. => W=W*
Wage adjustment:
• Expanding and contracting firms in both
• BUT: High-cost firms contract by more
• So: Demand for labour falls
• Wages fall, dampening merger incentives
Threshold sectors:
G (W , ~
z ;  )  Wa ( ~
z ) a  (1   )Wa * ( ~
z)  0



Labour-market equilibrium:
L  L(W , ~
z)

()
30
G (W , ~
z ;)  0
W



L  L(W , ~
z)

()
W0
L=L(.)
G(.; 0 )  0
½
~
z0
~
z
1
Fig. 3a: Simultaneous Determination of Wages and
Threshold Sectors: The No-Mergers Equilibrium
31
G(W , ~
z ;)  0
W



L  L(W , ~
z)

()
W0
W1
L=L(.)
W2
G(.; 0 )  0
G(.; 1 )  0
G(.; 2 )  0
½
~
z2
~
z1
~
z0
~
z
1
Fig. 3: Simultaneous Determination of Wages and
Threshold Sectors
32
c
F
O
H
HF
c*
Fig. 4: Wage Adjustments Dampen
Cross-Border Merger Waves
33
6. Mergers and Welfare
Partial-equilibrium intuition: W = CS + Sp
[Only changes in sectors where mergers occur matter]
1. Less competition  Some prices  CS 
2. High-cost firms are eliminated  Sp
 2. Dominates
[Lahiri-Ono EJ 1988]
GE: Consumers are also profit recipients
[Welfare is just the consumer’s (indirect) utility function]
1a. At given wages, only effect 1. matters  U 
2a. W  all prices   U
So: Full effect ambiguous in GE, but for different
reasons from partial equilibrium
34
Mergers and Welfare (cont.)
~
U  a 2  ( p ) 2
Indirect utility:
1
2
~
z
 p   p(W , z; n, n ) dz  ~z p(W , z;0, n* ) 2 dz
2
1
* 2
1
2
*
*
a

nW
a
(
z
)

n
W
a
( z)
*
 p(W , z; n, n ) 
n  n*  1
and
*
*
a

n
W
a
( z)
*
 p(W , z;0, n ) 
n*  1
At given wages, mergers (z~ ) raise prices in all sectors:
1
2
d p 2
* 2
* 2

p
(
W
,
z
;
n
,
n
)

p
(
W
,
z
;
0
,
n
)
~
dz
 p(W , z; n, n* )  p(W , z;0, n* )
b' n
 *
y (W , z; n, n* )
n 1
So:
~ ~
U  U (W , ~
z)

()
 0
IFF
y (.)  0
35
G(W , ~
z ;)  0

W


L  L(W , ~
z)

()
~
U  U (W , ~
z)

()
E0
~
U  U (.)
E1
L=L(.)
E2
G(.; 0 )  0
G(.; 1 )  0
G(.; 2 )  0
½
~
z
1
Fig. 5a: Mergers Increase Welfare
36
G(W , ~
z ;)  0

W


L  L(W , ~
z)

()
~
U  U (W , ~
z)

E1
()
E0
L=L(.)
~
U  U (.)
E2
G(.; 0 )  0
G(.; 1 )  0
G(.; 2 )  0
½
~
z
1
Fig. 5b: Mergers Reduce Welfare
37
Conclusion
Cross-border mergers are “instruments of comparative
advantage”:
• More specialisation
• Welfare may rise
Despite:
• Reduced competition in many sectors
• Redistribution from wages to profits
Empirical predictions:
• Trade liberalisation increases FDI
• Absent cost synergies, low-cost firms acquire highcost foreign rivals
• FDI & trade are complements
38
Conclusion
•
Cross-border M&As encouraged by trade liberalisation
39
40